Skip to content
Lenny's PodcastLenny's Podcast

Reganti & Badam: Why most AI products fail in production

Why treating LLMs as non-deterministic APIs and earning autonomy beats hype; human-in-the-loop calibration prevents the failures that sink AI products.

Lenny RachitskyhostAishwarya Naresh RegantiguestKiriti Badamguest
Jan 11, 20261h 26mWatch on YouTube ↗

CHAPTERS

  1. 0:00 – 5:03

    Introduction to Aishwarya and Kiriti

  2. 5:03 – 7:36

    Challenges in AI product development

  3. 7:36 – 13:19

    Key differences between AI and traditional software

  4. 13:19 – 15:23

    Building AI products: start small and scale

  5. 15:23 – 22:38

    The importance of human control in AI systems

  6. 22:38 – 25:18

    Avoiding prompt injection and jailbreaking

  7. 25:18 – 33:20

    Patterns for successful AI product development

  8. 33:20 – 41:27

    The debate on evals and production monitoring

  9. 41:27 – 45:41

    Codex team’s approach to evals and customer feedback

  10. 45:41 – 58:07

    Continuous calibration, continuous development (CC/CD) framework

  11. 58:07 – 1:01:24

    Emerging patterns and calibration

  12. 1:01:24 – 1:05:17

    Overhyped and under-hyped AI concepts

  13. 1:05:17 – 1:08:41

    The future of AI

  14. 1:08:41 – 1:14:04

    Skills and best practices for building AI products

  15. 1:14:04 – 1:26:22

    Lightning round and final thoughts

Get more out of YouTube videos.

High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.

Add to Chrome